This is practical. We've never seen black holes collide, ever. This is also gives us a powerful way to test general relativity in extreme gravity.
LIGO is limited to a certain frequency range, which I believe includes: stellar-mass black hole mergers (like this one), neutron star mergers, neutron star - black hole mergers, and (maybe) spinning neutron stars. There is also the Pulsar Timing Array that's been running for some time, that has its own frequency range. Finally there's the proposed eLISA observatory, which would be put in orbit of the Sun, and be able to measure the most massive black hole mergers, which we expect to happen some time after 2 galaxies merge. This is actually a key point to modern cosmology, which we have yet to observe happening. Our best evidence is "yup, don't see many BH pairs at the centers of galaxies, they must've merged already".
However, don't expect to use GW's to do most traditional astronomy tasks. They require two+ massive objects orbiting each other, or one asymmetric object spinning, so anything outside of that is off the table. There is the possibility, for some events, to observe both in GW's and with light, which would let us determine more about the system. Gamma ray bursts are a good example, though they're so short it's hard to point a traditional telescope in time.